Poster GC51C-1202, 45th Fall Meeting of the American Geophysical Union, December 3–7, 2012, San Francisco, CA, U.S.A.
NEON’s eddy-covariance flux data products – development of an explicit quality assurance and quality control framework Authors: Stefan R.I. Metzger1,2, Jeffrey R. Taylor1,2, Hongyan Luo1,2, and Henry W. Loescher1,2 (1) National Ecological Observatory Network, Boulder, CO, U.S.A.; (2) Institute for Arctic and Alpine Research, University of Colorado, Boulder, CO, U.S.A.
Background
The NEON flux QA/QC plan
Author: Leslie Goldman The National Ecological Observatory Network (NEON) is a continental-scale research platform with a projected operation of 30 years. NEON’s purpose is to provide high quality data products that will facilitate discovering and understanding the impacts of climate change, land-use change, and invasive species on ecology. The eddycovariance (EC) technique will be used to continuously monitor the exchange of sensible heat, water vapor, CO2 and other scalars between ecosystems and the atmosphere at all 60 NEON research sites.
Map of 20 NEON domains with indicators for different sites.
Why flux QA/QC? All data streams collected by NEON pass through an automated quality control (see companion poster GC51C1204). However, the EC method makes use of additional assumptions and simplifications, many of which are related to the mass balance equation; ௭
௭
I
ᇱ ᇱ
II ௭
௭
III ௭
V
௭
ᇱ ᇱ
ᇱ ᇱ
Four stages of scalable tests, resulting in Quality metrics (QM) Stage 1 “Temporally explicit flux QA/QC” • • • • • • •
Stationarity → QMSTA Developed turbulence → QMITC Comparison to reference spectrum → QMSPE Comparison to reference cospectrum → QMCOS Random error → QMERR Systematic error → QMERS Energy balance residual → QMRES
Stage 2 “Spatially explicit flux QA/QC” • Footprint modeling considering 3D dispersion
• Detailed case studies for selected sites • Analysis of Wavelet and Hilbert spectra for intermittent turbulence, coherent structures and gravity waves • Large eddy simulation to study advection, drainage flow, non-propagating eddies, and more realistic footprints
•
Enable the transition of EC measurements from principalinvestigator-based into observatory-based operations
•
Place QA/QC approaches into a production framework
•
Ensure consistent quality rating over the range of climates and ecosystems across an entire continent
•
Advance established tests and data flows (e.g., AmeriFlux, CarboEurope) to new state-ofthe-art functionality
•
Respond to requests from the research community
•
Provide research community with open source algorithms
• Superimpose footprints over high-resolution remote sensing data from NEON Airborne Observation Platform • Evaluation of target land cover in footprint → QMSOU
Criteria Eq. Term QMSTA QMITC QMSPE QMCOS QMERR QMERS QMRES QMSOU QMPUR Storage I Divergence II–III Divergence IV Advection V–VII Accuracy Random error Flow distortion Representativeness Stage 4 “Complex terrain”
Objectives
Stage 3 “Flux un-mixing in heterogeneous terrain” • High temporal resolution of fluxes from wavelet analysis, empirical mode decomposition or structure parameter • Footprint weights for individual land covers • Land cover specific fluxes from environmental response functions using boosted regression trees → QMPUR
IV
௭
VI
VII
Example I – Wavelet analysis (aircraft data*)
Example II – Footprint analysis (aircraft data*)
Example III – Environmental response function
with the total flux F into or out of an ecosystem, a scalar quantity s such as H2O or CO2 mixing ratios, along-, cross-, and vertical wind speeds u, v, and w with respect to the Cartesian coordinates x, y, and z; t is time, and z is the measurement height. Testing these assumptions and simplifications in addition to an automated quality control is indispensable to ensure high data quality as well as representativeness of the EC flux data products for the target ecosystems.
Contact Information:
[email protected]
www.neoninc.org
• Parseval’s theorem: covariance of a signal can be studied equivalently in time- and frequency domain • Integration over all turbulent scales for subsets of flux observations enables high temporal discretization
• Superimposing footprints over land cover maps • Fractional land cover contribution to each individual flux measurement
• Environmental response function (LTFM) relates sensible heat flux (response) to environmental drivers • Excellent agreement between observations and LTFM
*Courtesy of MAGIM - Matter fluxes in grasslands of Inner Mongolia as influenced by stocking rate (www.magim.net). The flight campaign was funded by the German Research Foundation, Research Group 536 MAGIM. Stipend funding by the German Academic Exchange Service, Helmholtz Association of German Research Centers, China Scholarship Council and the European Union under the Science and Technology Fellowship China is acknowledged. © National Ecological Observatory Network, Inc. ALL RIGHTS RESERVED. The National Ecological Observatory Network is a project solely sponsored by the National Science Foundation and managed under cooperative agreement by NEON, Inc. This material is based upon work supported by the National Science Foundation under the following grants: EF-1029808, EF-1138160, EF-1150319 and DBI-0752017. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
